Method for carrying out chess and card game behavior cloning based on factorization machine

A chess and card game behavior cloning method based on a factorization machine comprises the steps: selecting discrete input features as input, and expressing the discrete input features through a binary state that whether neurons are activated or not; and training a factorization machine model, sorting by changing the model and recalculating a loss function, selecting m groups of input feature pairs for combination to obtain newly added m input features, repeating until the number of neurons is greater than a set value, and connecting the finally combined input features with a model for retraining. A factorization machine is used for training to obtain the strength of the influence of correlation between neurons on a result to determine a connection structure when the neural network is constructed, a proper feature combination is extracted before a full connection layer, the characteristic that a common artificial neural network is convenient to calculate is reserved, the combination of multiple input featur